LFS-Bootscripts 2.2.0-pre1 released

Jeremy Utley jeremy at jutley.org
Sun Jul 11 09:01:59 PDT 2004


On Sun, 2004-07-11 at 08:34, Archaic wrote:
> On Sun, Jul 11, 2004 at 01:22:20PM +0200, Matthias Benkmann wrote:
> > 
> > How would single-user mode help on remote systems?
> 
> It wouldn't. It is just another thing I keep on my mind because I run
> multiple remote systems.
> 
> > But all of this is way beyond basic LFS.
> 
> If a clean method can be found that covers more bases, then it
> absolutely should be put it. However, I am not advocating something
> overly specialized that would benefit only one segment of users. In
> fact, I haven't advocated anything. What I thought about when I made my
> first reply in this thread was that if a user simply forgot to add the
> ramfs support in a new kernel he was testing, it would be trivial enough
> to fix in single user mode avoiding the need for a host or boot disk.
> Also, is there any static device backup on these udev systems? If not,
> then one would have to have a specialized boot disk. Not that that isn't
> a bad idea to have, but still seems overkill for a simple kernel
> rebuild.

As long as your running a 2.6.x kernel (pretty much a requirement for
any system covered by LFS 6), ramfs is available, and unable to be
disabled.  As far as static backup, IIRC 2 device nodes (console and one
other) are created on the underlying filesystem in the book - this
should be enough to at least get to a command prompt, at which point
mknod *COULD* be used if needed.

-J-




More information about the lfs-dev mailing list